This Isn't Going To Work

Or we could build technology that doesn't cause people to crash cars when using it. Such technology would also have the benefit of being nicer and easier to use than looking down at a phone while driving.

Trying to lock users out is doomed to fail. Many solutions would have a "snooze button" effect where people might be more distracted by the additional effort to get around the lock. Others might end up being far too restrictive (what about passengers? people in buses or trains? people in emergency situations?).

Technology is also only a small part of distracted driving. It's still perfectly legal for me to eat a Big Mac and fries while having a heated debate with passengers and adjusting the car stereo while driving. We ought to build (1) a culture that strongly discourages this, and (2) cars that build in safety features (driver alertness detection, etc) while reducing the attention requirements of the driver (automatic collision avoidance, self-driving cars...)

Technical solutions to cultural problems often seem to result in arms races of various sorts. In part this has to do with the natural difference between actual and perceived problems and the constraints of technology.

"[W]hen a phone is moving, don't permit it to accomplish certain tasks." That sounds like a simple and efective enough solution until you're a passenger. Then it becomes a nuisance, some pain point to be hacked around.

sensors in an auto could detect the position of each phone relatively to the steering wheel and only disable the ones that could belong to the driver.

or, said the other way, enable the passengers to override the automatic disablement if they request and their position is determined to be passenger.

there will still be hacks but if we could take care of x% of the problem then we would be far better off then today.